CRNA Opt-Out States: The Complete List, and What Opt-Out Actually Means

On July 2, 2026, Vermont Governor Phil Scott sent a letter to the Centers for Medicare & Medicaid Services opting Vermont out of the federal physician supervision requirement for CRNAs. Vermont is the 26th state to do it, and the third in New England. More than half the country has now opted out.

Below is the complete list with dates and scope. Then, because the term causes more confusion than almost anything else in CRNA practice, an explanation of what opting out actually changes — and the several things it very much does not.

Every CRNA opt-out state

Guam opted out in June 2016, but as a territory it doesn’t appear in the state count. Dates are from AANA’s opt-out fact sheet; the ASA maintains its own tracker of the same states from the opposing side of the policy debate, which makes a useful cross-check.

Two of the 26 are partial

Most lists you’ll find report a flat count. That flattens a distinction that matters if you’re making decisions on it.

Utah (2022) and Wyoming (2023) opted out only for critical access hospitals and specified small rural hospitals — Wyoming’s is limited to hospitals with 25 or fewer licensed beds. Colorado was likewise partial from 2010 until it obtained a full opt-out in October 2023.

So the honest count is 24 full opt-outs and 2 partial ones. Montana’s has also been reversed and reinstated: its governor withdrew the 2004 opt-out in May 2005 and restored it that June.

What opt-out actually is

The federal supervision requirement lives in three places in the Code of Federal Regulations. Each is a condition a facility must satisfy to participate in, or be covered by, Medicare:

Read who the supervisor is allowed to be. Under 482.52(a), anesthesia may be administered by a CRNA “under the supervision of the operating practitioner or of an anesthesiologist who is immediately available if needed.” The operating practitioner. The surgeon already scrubbed in satisfies the federal rule. It has never required an anesthesiologist.

That single detail undercuts most of what people assume the rule does — and it raises the obvious question of why opt-out matters at all, which we’ll come back to.

How a state opts out

CMS created the opt-out in a final rule published November 13, 2001. To use it, a state’s governor sends CMS a letter attesting to three things:

1. The governor consulted the state’s boards of medicine and nursing about access to and quality of anesthesia services;
2. Opting out is in the best interests of the state’s citizens; and
3. The opt-out is consistent with state law.

Governor Scott’s letter tracks that template. He wrote that he had “consulted with Vermont’s Board of Medical Practice, Vermont Board of Nursing, and Office of Professional Regulation regarding issues related to access to and quality of anesthesia services in Vermont,” and had “concluded it’s in the best interests of Vermonters to opt-out of the current physician supervision requirements.”

Two features of this process explain why opt-out is a weaker instrument than its reputation suggests.

It is effective on submission. No CMS review, no approval, no waiting period. Per the rule, a governor’s letter “will be accepted on face value, with no independent CMS scrutiny or analysis.” CMS declined to define “consultation,” declined to set criteria for “best interests,” and declined to require so much as an attorney general’s opinion on consistency with state law. A governor can also withdraw an opt-out at any time, equally effective on submission — which is how Montana’s was reversed and restored inside of two months.

The third attestation is the tell. The governor must certify the opt-out is consistent with state law, meaning state law already had to permit unsupervised practice. Opt-out cannot manufacture that authority; it can only acknowledge authority that was already there. By AANA’s own count, 45 states have no physician supervision requirement for CRNAs in their nursing or medical laws and regulations at all. Those states were never waiting on CMS.

What opt-out does not do

It does not change state scope of practice. Scope comes from the state practice act and board rules. Unchanged.

It does not change the professional fee. The anesthesia professional service is billed under Medicare Part B, by the provider. Opt-out reaches facility participation and coverage conditions and never touches the professional fee. Be careful with the common shorthand that opt-out “only affects the Part A facility fee” — it’s close, but wrong. Hospital and CAH participation is the Part A side, but ASC facility fees are paid under Part B, and the ASC condition at 416.42 is squarely within the opt-out. The precise statement: opt-out governs facility conditions, not the professional fee.

It does not override facility policy. Hospital bylaws, credentialing, and delineation of privileges still control who does what. A hospital in an opt-out state remains entirely free to require anesthesiologist supervision, and many do.

It does not move liability. Liability is allocated by state tort law, employment relationships, and insurance — not by a Medicare condition of participation.

And in a state whose law already permitted unsupervised practice, it does not by itself change anything clinically. The anesthetic delivered on Wednesday looks like the one delivered on Monday.

So why does it matter?

Two reasons, and the first is the one that gets left out of the explainers.

Supervision requires a willing supervisor

The regulation permits the operating surgeon to supervise. But supervision isn’t a fact about who is standing in the room — it’s a role someone has to accept. And surgeons are frequently unwilling to accept it, because they believe being named the supervising physician exposes them to liability for anesthesia decisions they did not make.

Are they right? Largely not. Courts have moved away from the “captain of the ship” doctrine, though it varies by state, and liability generally turns on the right to control the CRNA’s actions rather than on holding a supervisory title. AANA maintains standing guidance on surgeon liability and its general counsel has written at length on what supervision does and does not mean. Organizations don’t build that material for a fear nobody holds.

The perception persists, and perception is sufficient. A surgeon who won’t sign, a malpractice carrier that advises against it, or a hospital counsel unwilling to take the chance — any one of them can veto a CRNA-only model. The federal rule hands each of them the veto by using the word.

Opt-out doesn’t grant CRNAs authority they lacked. It removes a veto.

The stakes at the facility level are disproportionate

The supervision requirement is a condition of participation. A deficiency there doesn’t produce a fine — it puts the facility’s Medicare participation at risk. That is an existential-scale consequence attached to a staffing-model question, which is precisely why administrators are conservative about it.

It also explains where opt-out bites hardest. CRNAs are the predominant anesthesia providers in rural America, and many rural facilities are critical access hospitals with no anesthesiologist on staff. For them the question was never which model is better; it was whether they can offer surgical services and be paid for them. That’s what Vermont’s governor pointed at, and it’s why Utah’s and Wyoming’s partial opt-outs are scoped specifically to critical access and small rural hospitals.

The perception problem, plainly

Supervision is a load-bearing word carrying weight it was never designed to carry.

Hospital counsel reads it and hears liability exposure. Surgeons read it and infer they are answerable for anesthesia decisions they are not making. Administrators read it and see regulatory risk. None of those inferences follow from 42 CFR 482.52 — the rule is a participation condition, satisfied by the surgeon’s presence — but the word invites them.

That ambiguity is usable. When a facility weighs a medically-directed care team against a CRNA-only model, “federal law requires supervision” sounds like a legal constraint while actually describing a Medicare billing condition the surgeon in the room already satisfies. Removing the word removes the argument.

The outcomes question, stated fairly

You will see it asserted that the safety data on unsupervised CRNA practice is settled. The research is more substantial than any single study, and the two professional societies read it differently.

The best-known paper is Dulisse and Cromwell, Health Affairs, 2010, which examined Medicare data from 1999–2005 across roughly 481,000 hospitalizations and found no evidence that opting out produced increased inpatient deaths or complications. AANA points to it and to subsequent work; the ASA disputes the interpretation on methodological grounds. If you want to reason about this yourself, read both societies’ positions and the underlying papers rather than either side’s summary of the other.

One thing both sides can agree happened: when CMS created the opt-out in 2001, it said the Agency for Healthcare Research and Quality would study anesthesia outcomes in opt-out states versus non-opt-out states. Per AANA’s most recent fact sheet, that specific study was never initiated. Twenty-five years on, the agency that wrote the requirement never produced the comparison it promised. That is a statement about AHRQ, not about the literature, which has continued without it.

What this means if you’re evaluating a job

Opt-out status tells you remarkably little about any specific position.

Facility bylaws govern your day. There are hospitals in states that opted out in 2002 that still require anesthesiologist supervision by internal policy, and there are non-opt-out states where the surgeon signs without a second thought. The governor’s letter determines what a facility is allowed to do. It does not determine what it does.

When you’re evaluating a role — in Vermont or anywhere on the list above — the questions that decide your practice are what the anesthesia staffing model actually is, what the medical staff bylaws require, and how anesthesia privileges are delineated. Ask those three directly. They matter more than the map.

Sources

• Vermont Opts Out of Physician Supervision of CRNAs — AANA press release, July 9, 2026, quoting Gov. Scott’s July 2, 2026 letter.
• Fact Sheet Concerning State Opt-Outs and the November 13, 2001 CMS Rule — AANA. Source for the state list, dates, partial opt-outs, the three attestations, and CMS’s rule commentary.
• Opt-Outs — American Society of Anesthesiologists. The same list, maintained by the opposing advocacy organization.
• Medicare and Medicaid Programs; Hospital Conditions of Participation: Anesthesia Services — the November 13, 2001 CMS final rule creating the opt-out.
• 42 CFR 482.52 — Hospital Condition of Participation, Anesthesia Services.
• 42 CFR 485.639 — Critical Access Hospital Condition of Participation, Surgical Services.
• 42 CFR 416.42 — Ambulatory Surgical Center Condition for Coverage.
• Surgeon Liability and The Nature of Supervision — AANA guidance on what supervision does and does not create.
• Dulisse B, Cromwell J. No Harm Found When Nurse Anesthetists Work Without Supervision By Physicians. Health Affairs. 2010;29(8):1469–1475.
• ASA Statement Regarding AANA-Sponsored Paper Published in Health Affairs — August 2010 response.